What is this matrix decomposition, and is it useful for solving least square problems?

69 Views Asked by At

I am numerically studying a class of matrices $A$, size $n \times n$, that are singular, but can be decomposed into $$A = V D V^T,$$ where $D$ is a (non-singular) square, diagonal matrix of size $m \times m$ and $V$ is a matrix of size $n \times m$. Does this decomposition have a name?

I am trying to solve for $$ \arg \min_x ||A x - b||^2.$$ My question is: is the known decomposition useful in any way? Had it been an SVD decomposition, it would immediately give the answer to the equation.

In general $n < m$, and an example of $V$ could be

V = [[-1, -1,  0, -1,  0,  0,  0,  0],
     [ 1,  0, -1,  0,  0, -1,  0,  0],
     [ 0,  1,  1,  0, -1,  0, -1,  0],
     [ 0,  0,  0,  1,  1,  0,  0, -1],
     [ 0,  0,  0,  0,  0,  1,  1,  1]]

$D$ can then be any diagonal matrix, but with the identity matrix, we have

A = [[ 3, -1, -1, -1,  0],
     [-1,  3, -1,  0, -1],
     [-1, -1,  4, -1, -1],
     [-1,  0, -1,  3, -1],
     [ 0, -1, -1, -1,  3]]

which has rank $n - 1$.